Transcriptome profiling, molecular biological, and physiological studies reveal a major role for ethylene in cotton fiber cell elongation

Abstract

Upland cotton (Gossypium hirsutum) produces the most widely used natural fibers, yet the regulatory mechanisms governing fiber cell elongation are not well understood. Through sequencing of a cotton fiber cDNA library and subsequent microarray analysis, we found that ethylene biosynthesis is one of the most significantly upregulated biochemical pathways during fiber elongation. The 1-Aminocyclopropane-1-Carboxylic Acid Oxidase1-3 (ACO1-3) genes responsible for ethylene production were expressed at significantly higher levels during this growth stage. The amount of ethylene released from cultured ovules correlated with ACO expression and the rate of fiber growth. Exogenously applied ethylene promoted robust fiber cell expansion, whereas its biosynthetic inhibitor l-(2-aminoethoxyvinyl)-glycine (AVG) specifically suppressed fiber growth. The brassinosteroid (BR) biosynthetic pathway was modestly upregulated during this growth stage, and treatment with BR or its biosynthetic inhibitor brassinazole (BRZ) also promoted or inhibited, respectively, fiber growth. However, the effect of ethylene treatment was much stronger than that of BR, and the inhibitory effect of BRZ on fiber cells could be overcome by ethylene, but the AVG effect was much less reversed by BR. These results indicate that ethylene plays a major role in promoting cotton fiber elongation. Furthermore, ethylene may promote cell elongation by increasing the expression of sucrose synthase, tubulin, and expansin genes.

Figure 1.

TreeView Representation of Fiber-Specific Cotton ESTs and Analysis of Data Quality. (A) Top panel: hierarchical clustering of 2522 ESTs that showed FDR-corrected P values <0.001 in at least one of the growth stages. The signals are shown in a red-green color scale, where red represents higher expression and green represents lower expression. The numbers represent the DPA of ovule harvest of the hybridizing RNA. An RNA sample from 3-DPA ovules was used as the reference for each hybridization. a and b, genes induced before or after 3 DPA and maintained at relatively high levels throughout the experimental period; c, genes induced before 3 DPA and repressed drastically around 10 DPA. Bottom panel: hierarchical clustering of 778 ESTs that were developmentally upregulated in wild-type ovules but not in the mutant. a1, genes induced at 3 DPA with peak levels found at 5 to 10 DPA; a2, genes induced at 3 DPA and peaking around 10 to 20 DPA; b1, genes induced at 5 DPA and peaking around 10 to 20 DPA; b2, genes induced at 5 DPA with peak levels found at 5 to 10 DPA; c1, genes repressed at 15 DPA; c2, genes repressed at 5 or 10 DPA. (B) Experimental variation and reproducibility assessment from randomly chosen microarray hybridizations. Top panel: comparisons of expression ratios obtained from swap-dye experiments to show the labeling efficiency of different dyes. Bottom panel: self-hybridization results obtained after probing the microarray with the same RNA sample prepared from 3-DPA wild-type ovules and labeled separately with either Cy3 or Cy5 dye. (C) Scatterplot comparisons of 10/3-DPA hybridization data showing systematic upregulation of a large fraction of ESTs during the fast cell elongation period.

Figure 2.

Detailed Expression Profiles of Genes in Major Plant Hormone Biosynthetic Pathways That Showed FDR-Corrected P Values <0.001. (A) to (D) Comparison of expression ratios obtained from six microarray hybridizations for genes involved in ethylene, BR, GA, and auxin biosynthesis. Bottom panels of (A) and (B): For data verification, QRT-PCR analysis was performed on ACO1-3, SMT1, and DET2, which were regarded as fiber-preferentially expressed genes after analysis of microarray hybridization data with FDR-corrected P values <0.001. Relative expression levels were determined after normalizing all data to that of 3-DPA wild-type ovules, which was set to 1.0. Error bars represent sd for three independent experiments. The time (DPA) of ovule collection is indicated. ACS6, ACC synthase 6; ACO, ACC oxidase; SMT1, 24-sterol C-methyltransferase; DEM1, steroid demethylase; DWF5, sterol δ7 reductase; DWARF1, C-24 sterol reductase; DET2, steroid 5-α-reductase; DDWF1, putative cytochrome P450 gene catalyzing typhasterol to castasterone; GA20ox, GA 20-oxidase; GA3ox, GA 3-hydroxylase; NIT, nitrilase; FMO1, flavin-containing monooxygenase; TDC, tryptophan decarboxylase.

Figure 3.

Tissue-Specific Expression, Enzymatic Activity of ACO1-3 in Yeast, and Ethylene Production from Different Cotton Samples. (A) ACO1-3 transcripts are specifically expressed in developing cotton fibers. ACO1-3 expression levels are quantified by QRT-PCR, and data of all samples are normalized to that of 10-DPA wild-type ovules with fiber cells removed, which is set to 1.0. Error bars represent sd for three independent experiments. F, fibers only; F+O, fibers attached on the ovules; O, ovules with fibers removed. (B) Enzyme activity in yeast cells expressing each of the ACO genes. Air samples (100 μL) from each reaction containing 10⁶ yeast cells, cofactors, and various amounts of the substrates were removed and injected into the column of a gas chromatograph for ethylene measurements. Means ± sd (bars) obtained for each substrate concentration point were calculated from measurements of three different cultures of the same transformant cell line with triplicate measurements of each cell culture. Closed symbols indicate results obtained from galactose-induced cells, and open symbols indicate that of noninduced cells. (C) Only wild-type ovules produced substantial amounts of ethylene. Mutant or wild-type ovules (∼30 collected from two flowers) were cultured for a total of 12 d, and ethylene production was analyzed in a gas chromatograph by injecting 100 μL of the head air from the culture flasks (with total available volume of ∼50 mL) directly to the HP-PLOTQ column. Wild-type ovules were also cultured in the presence of 1.0 μM AVG for ethylene measurements. Means ± sd (bars) were calculated from measurements of three different ovule cultures (or closed glass jars for seedlings) with triplicate measurements for each sample. FW, fresh weight.

Figure 4.

Exogenous Ethylene Promotes and AVG Inhibits Fiber Cell Elongation. (A) Phenotypes of 7-d-old wild-type ovules (collected at 1 DPA) cultured with or without (CK) ethylene supplementation. Bar = 2.5 mm. (B) Final fiber lengths measured at the end of the 6-d culture period. (C) Ovule sizes measured at the end of the 6-d culture period. (D) Phenotypes of 13-d-old wild-type ovules (collected at 1 DPA) cultured with or without (CK) AVG. Bar = 5 mm. (E) Final fiber lengths measured at the end of the 12-d culture period. (F) Ovule sizes measured at the end of the 12-d culture period. Ovules were cultured for a longer period in (D) to (F) to maximize the differences in fiber length between the AVG-treated and the non-AVG-treated ovules. Each data point in (B) and (E) is the average of three independent ovule culture experiments, with a total of 90 fiber cells measured on three individual ovules every time. Each data point in (C) and (F) is the average of 30 ovules obtained from three independent culture experiments. Error bars indicate sd (n = 30).

Figure 5.

Effects of Exogenously Applied BR and BRZ on Fiber Cell Elongation. (A) Phenotypes of 7-d-old ovules (collected at 1 DPA) cultured with and without (CK) BR supplementation. Bar = 2.5 mm. (B) Final fiber lengths measured at the end of the 6-d culture period. (C) Ovule sizes measured at the end of the 6-d culture period. (D) Phenotypes of 13-d-old ovules (collected at 1 DPA) cultured with or without (CK) BRZ. Bar = 5 mm. (E) Final fiber lengths measured at the end of the 12-d culture period. (F) Ovule sizes measured at the end of the 12-d culture period. See legend of Figure 4 for details on sample preparation and measurements. Error bars indicate sd (n = 30).

Figure 6.

Ethylene and BR Interact during Fiber Elongation, Ovule Cell Expansion, and Gene Expression. (A) Final fiber lengths measured at the end of a 6-d culture period with ethylene, BR, or their biosynthetic inhibitors added individually or in combination. Wild-type ovules collected at 1 DPA were used throughout the study. Ethylene (0.1 μM), BR, and the two inhibitors (1 μM) were added in the culture as defined in the figure. CK, no chemicals added. (B) Ovule sizes measured at the end of a 6-d culture period with ethylene, BR, or their biosynthetic inhibitors added individually or in combination. See legend of Figure 4 for details on sample preparation and measurements. Error bars indicate sd (n = 30). (C) The effect of ethylene on BR biosynthetic gene expression. (D) The effect of BR on ethylene biosynthetic gene expression. Total RNA samples prepared from 1-DPA ovules after culturing for the specified time (h) in the presence of 5 μM ethylene (ET) or 5 μM BR were used for RT-PCR analysis. Data are representative of three amplifications from independent RNA preparations. UBQ7 was used as a normalization control. Significantly upregulated genes upon respective hormone treatment are shown in bold in (C) and (D).

Figure 7.

Expression Profiling of Five Downstream Genes Important for Fiber Elongation in Wild-Type and fl Mutant Ovules after Ethylene or AVG Treatment. Total RNA samples prepared from 1-DPA ovules after a 3-d culture with or without adding 5 μM ethylene (ET) or 1 μM AVG were used for RT-PCR analysis. Data are representative of three amplifications from independent RNA preparations.